Cumulative Convolutional Codes for Communication Systems

Abstract

The present disclosure relates to an accumulative iterative code (AIC) for a communication system including a transmitter and a receiver. The technical solutions disclosed herein are based on the theory of interactive channel coding with point-to-point iterative channel coding for transmission over a channel with feedback based on iterative refinement. The iterative refinement includes transmitting an uncoded word in an initial transmission and then iteratively refining the information of the transmitted word that the receiver has through subsequent transmissions until the receiver is able to generate the transmitted word. Each subsequent transmission is based on the word and the feedback information obtained from the receiver through the feedback channel. The schemes based on iterative refinement are candidates most suitable for ultra-high reliability transmission because the error probability of these schemes decays doubly exponentially with the codeword length, i.e., significantly faster than the single exponential decay that may be achieved without feedback. In addition, the present disclosure also relates to a corresponding method and a computer program.

Description

Technical Field

[0001] The present disclosure relates to a first communication device and a second communication device for providing cumulative convolutional codes for a communication system. In addition, the present disclosure also relates to a corresponding method and a computer program. Background Art

[0002] The 3GPP fifth generation (5G) wireless network is expanding its connectivity and service provision capabilities to new vertical domains. In the field of industrial automation, there is an increasing interest in providing wireless connectivity to communication devices involved in supervising, assisting, and controlling production processes. The applications with the strictest reliability and timing constraints are the motion control of production machinery, i.e., motion control is expected to benefit from the introduction of wireless connectivity to replace the cable links used to connect production machinery to its control system.

[0003] The real-time control of production processes combines very high reliability requirements and stringent latency constraints. In real-time motion control applications, all exchanged messages must be properly protected, and the probability of two consecutive packet errors can be neglected. This is because a single packet error is tolerable, but two consecutive packet errors may damage the machine, resulting in production downtime and potentially huge economic losses. Communication service availability is a basic requirement for motion control systems. Summary of the Invention

[0004] An object of an example of the present disclosure is to provide a technical solution that reduces or solves the disadvantages and problems of traditional technical solutions.

[0005] Another object of an example of the present disclosure is to overcome the disadvantages of traditional technical solutions while reliably transmitting words / messages with a significant improvement in spectral efficiency.

[0006] The above and other objects are achieved by the subject matter of the independent claims. Other advantageous examples of the present disclosure can be found in the dependent claims.

[0007] According to a first aspect of the present disclosure, the above and other objects are achieved by a first communication device for a communication system. The first communication device is configured to:

[0008] transmit a first word x to a second communication device in a first channel 0 , where the first word x 0 is generated based on an information word m;

[0009] i) receive a second word p from the second communication device in a second channel i , where the second word p i indicates a first word x transmitted in the first channel iA set of associated bits and a set of corresponding reliability values where i is a time index,

[0010] ii) Determine an error word e based on the first transmitted word x i and the set of bits to determine an error word e i ,

[0011] iii) Compress the error word e based on the set of reliability values to generate a first word x i , and i+1 and

[0012] iv) Transmit the first word x to the second communication device in the first channel i+1 .

[0013] The set of bits and the set of corresponding reliability values may have a one-to-one mapping. The set of corresponding reliability values may be, for example, positive integers indicating different reliability values.

[0014] The words in this document may also represent messages.

[0015] The advantage of the first communication device provided according to the first aspect is that the error word is compressed based on the reliability value to obtain a compressed error word with the minimum length, thereby reducing the use of frequency and time resources of the first channel.

[0016] According to the first aspect, in an implementation of the first communication device, the first communication device is configured to:[[]]

[0017] Repeat i) to iv) D times, where i = 0,..., D - 1, and D depends on at least one stop condition.

[0018] The advantage of this implementation is that by repeating steps i) to iv), an iterative refinement process is completed, thereby increasing the possibility that the second communication device correctly decodes the received word.

[0019] According to the first aspect, in an implementation of the first communication device, the first stop condition is the error-free reception of the first word x D in the second communication device, and the first communication device is configured to:[[]]

[0020] Transmit a first control signal to the second communication device, where the first control signal indicates the error-free reception of the first word x D .

[0021] The advantage of this implementation is that once error-free reception occurs, the first control signal stops the iterative refinement process, thus avoiding the use of more frequency and time resources for iterative refinement.

[0022] According to the first aspect, in an implementation of the first communication device, the second stop condition is the delay requirement of the information word m.

[0023] The advantage of this implementation is that whenever the iterative refinement process is about to exceed a given delay (beyond which the transmission delay is unacceptable), the iterative refinement process stops.

[0024] According to the first aspect, in an implementation of the first communication device, the error word e i indicates that the first word x i and the set of bits have different values.

[0025] The advantage of this implementation is that the error word precisely indicates which bits of the previously transmitted data need to be corrected, enabling error correction in the second communication device.

[0026] According to the first aspect, in an implementation of the first communication device, the set of bits and the set of corresponding reliability values are determined based on a set of log-likelihood ratio (LLR) z i determined.

[0027] The advantage of this implementation is that determining a set of bits and a set of corresponding reliability values based on LLR can be used with any modulation, as it is well-known that bit LLR can be calculated for any modulation.

[0028] According to the first aspect, in an implementation of the first communication device, generating the first word x i+1 includes:

[0029] Adding bits to the first word x transmitted i to generate the first word x i+1 , where the reliability value of the first word x transmitted i is 0.

[0030] The advantage of this implementation is that bits with low reliability, such as those less than a given threshold or level, are retransmitted without compression, as compression does not effectively reduce the size of the corresponding error word. The value of the above threshold or level can be adapted to different applications.

[0031] According to the first aspect, in an implementation of the first communication device, the first communication device is used for:

[0032] Add at least one redundant bit to make the number of bits of the first word x i+1 adapt to the frequency and time resources for transmission.

[0033] The advantage of this implementation is that the redundant bits can be used for error detection / correction in the second communication device, thereby further improving the transmission reliability.

[0034] According to a first aspect, in an implementation of a first communication device, the first communication device is configured to:

[0035] Receive a second control signal from the second communication device, where the second control signal indicates the length of the second word p i ;

[0036] Determine a generator matrix G based on the length of the second word p i ; and i ; and

[0037] Based on the generator matrix G i and the second word p i , obtain a set of bits and a corresponding set of reliability values

[0038] The advantage of this implementation is that errors in the corrupted second word can be corrected based on the generator matrix.

[0039] According to a second aspect of the present disclosure, the above and other objects are achieved by a second communication device for a communication system. The second communication device is configured to:

[0040] Receive a first word x from the first communication device in a first channel 0 , where the first word x 0 is generated based on an information word m;

[0041] i) Determine a set of bits i associated with receiving the first word x in the first channel and a corresponding set of reliability values where i is a time index;

[0042] ii) Transmit a second word p to the first communication device (100) in a second channel i , where the second word p i indicates the set of bits and the corresponding set of reliability values

[0043] iii) Receive the first word x from the first communication device in the first channel i+1 .

[0044] An advantage of the second communication device provided by the second aspect is that transmitting the second word enables the error word to be calculated in the first communication device.

[0045] According to the second aspect, in one implementation of the second communication device, the second communication device is configured to:

[0046] Repeat i) to iii) D times to obtain multiple groups of bits and multiple groups of corresponding reliability values where i = 0, …, D - 1 and D depends on at least one stop condition.

[0047] An advantage of this implementation is that by repeating steps i) to iii), an iterative refinement process is completed, thereby increasing the likelihood of the second communication device correctly decoding the received word.

[0048] According to the second aspect, in one implementation of the second communication device, the first stop condition is the error - free reception of the first word x D in the second communication device, and the second communication device is configured to:

[0049] Receive a first control signal from the first communication device, where the first control signal indicates the error - free reception of the first word x D in the second communication device.

[0050] An advantage of this implementation is that once error - free reception occurs, the first control signal stops the iterative refinement process, thereby avoiding using more forward / channel frequency and time resources for iterative refinement.

[0051] According to the second aspect, in one implementation of the second communication device, the second stop condition is the latency requirement of the information word m.

[0052] An advantage of this implementation is that whenever the iterative refinement process is about to exceed a given latency (exceeding the given latency, the latency is unacceptable), the iterative refinement process stops.

[0053] According to the second aspect, in one implementation of the second communication device, the group of bits and the group of corresponding reliability values are determined based on a group of LLRz i and determined.

[0054] An advantage of this implementation is that determining a group of bits and a group of corresponding reliability values based on LLR can use any modulation because it is well - known that bit LLR can be calculated for any modulation.

[0055] According to a second aspect, in an implementation of the second communication device, the second communication device is configured to:

[0056] decode the multiple groups of bits and the multiple groups of corresponding reliability values to obtain the information word m.

[0057] According to a second aspect, in an implementation of the second communication device, decoding the multiple groups of bits and the multiple groups of corresponding reliability values includes:

[0058] i) Based on the first word x i and the group of corresponding reliability values correct errors in a group of bits to obtain the first word x i-1 ;

[0059] Repeat i) D times to obtain the first word x 0 ; and

[0060] Based on the first word x 0 determine the information word m.

[0061] The advantage of this implementation is that the second communication device determines the information word.

[0062] According to a second aspect, in an implementation of the second communication device, the second communication device is configured to:

[0063] Based on the length of the first word x i select the generator matrix G i ;

[0064] Based on the group of bits the group of corresponding reliability values and the generator matrix G i obtain the second word p i ; and

[0065] Transmit a second control signal to the first communication device, where the second control signal indicates the length of the second word p i .

[0066] The advantage of this implementation is that the second communication device obtains a shorter second word by using the generator matrix, thereby reducing the use of time-frequency resources of the second channel.

[0067] According to a third aspect of the present disclosure, the above and other objects are achieved by a method for a first communication device. The method includes:

[0068] Transmit a first word x to a second communication device in a first channel 0 , wherein the first word x 0 is generated based on an information word m;

[0069] Receive a second word p from the second communication device in a second channel i , wherein the second word p i indicates a set of bits i associated with the first word x transmitted in the first channel and a set of corresponding reliability values wherein i is a time index,

[0070] Determine an error word e based on the transmitted first word x i and the set of bits , i ,

[0071] Compress the error word e based on the set of reliability values to generate a first word x i , and i+1 ,

[0072] Transmit the first word x to the second communication device in the first channel i+1 .

[0073] The method provided by the third aspect can be extended to an implementation corresponding to the implementation of the first communication device provided by the first aspect. Therefore, an implementation of the method includes one or more features of the corresponding implementation of the first communication device.

[0074] The advantages of the method provided by the third aspect are the same as those of the corresponding implementation of the first communication device provided by the first aspect.

[0075] According to a fourth aspect of the present disclosure, the above and other objects are achieved by a method for a second communication device. The method includes:

[0076] Receive a first word x from a first communication device in a first channel 0 , wherein the first word x 0 is generated based on an information word m;

[0077] Determine a set of bits i associated with receiving the first word x in the first channel and a set of corresponding reliability values wherein i is a time index;

[0078] Transmit a second word p to the first communication device in a second channel i, wherein the second character p i indicates the set of bits and the corresponding set of reliability values

[0079] receive a first word x from the first communication device in the first channel i+1 .

[0080] The method provided in the fourth aspect can be extended to an implementation corresponding to the implementation of the second communication device provided in the second aspect. Therefore, an implementation of the method includes one or more features of the corresponding implementation of the second communication device.

[0081] The advantages of the method provided in the fourth aspect are the same as those of the corresponding implementation of the second communication device provided in the second aspect.

[0082] The present disclosure also relates to a computer program having program code. When the program code is run by at least one processor, the at least one processor is caused to execute any method provided by the examples of the present disclosure. In addition, the present disclosure also relates to a computer program product, including a computer-readable medium and the computer program, wherein the computer program is included in the computer-readable medium and includes one or more of the following groups: Read-Only Memory (ROM), Programmable ROM (PROM), Erasable PROM (EPROM), flash memory, Electrically EPROM (EEPROM), and hard disk drive.

[0083] Other applications and advantages of the examples of the present disclosure will be apparent from the following detailed description. BRIEF DESCRIPTION OF THE DRAWINGS

[0084] The drawings are intended to illustrate and explain different examples of the present disclosure, wherein:

[0085] Figure 1 shows a first communication device provided by an example of the present disclosure;

[0086] Figure 2 shows a method for a first communication device provided by an example of the present disclosure;

[0087] Figure 3 shows a second communication device provided by an example of the present disclosure;

[0088] Figure 4 shows a method for a second communication device provided by an example of the present disclosure;

[0089] Figure 5 shows the interaction between a first communication device and a second communication device in a wireless communication system provided by an example of the present disclosure;

[0090] Figure 6 shows the iterative interaction between a first communication device and a second communication device in a communication system provided by another example of the present disclosure;

[0091] Figure 7 shows a second communication device provided by an example of the present disclosure;

[0092] Figure 8a and Figure 8b respectively show a first communication device and a second communication device provided by an example of the present disclosure;

[0093] Figure 9 shows the use of a generator matrix provided by an example of the present disclosure;

[0094] Figure 10 shows the iterative interaction between a first communication device and a second communication device in a communication system provided by yet another example of the present disclosure; and

[0095] Figure 11 shows the performance results of an example of the present disclosure. Detailed Description

[0096] In traditional communication systems, high reliability is typically achieved through Error Correction Coding (ECC). ECC requires the transmission of long codewords to achieve high reliability without significantly reducing the data rate. When the information word to be transmitted is very short, using long codewords will obviously reduce the data rate. Even when using state-of-the-art ECC techniques, such as 3GPP New Radio (NR) polar codes, a large part of the data rate must be sacrificed to obtain high reliability for short words. For low SNR (e.g., about 0 dB or below), the reduction in data rate caused by short-word transmission is even more obvious, where the data rate achievable with short words is less than half of the data rate achievable when transmitting long words.

[0097] Therefore, different traditional technical solutions with iterative coding have been proposed. However, traditional technical solutions have many drawbacks that make their implementation impractical or even impossible, for example, a noiseless feedback channel / an unlimited feedback data rate; it is not clear how to perform traditional modulation; based on complex coding schemes, i.e., convolutional codes with a large number of states, non-binary LDPC codes; a large number of iterations, depending on the word length.

[0098] Thus, according to examples of the present disclosure, an Accumulative Iterative Code (AIC) is disclosed herein to very reliably transmit messages / words at high spectral efficiency. Examples of the present disclosure are applicable to short information words and long information words. A short information word may, for example, include less than 1000 information bits. In some application areas, the typical length of a word is between 20 bytes and 50 bytes, i.e., between 160 bits and 400 bits. For example, 3GPP Long Term Evolution (LTE) and Fifth Generation (5G) New Radio (NR) specify the coding of information words with a length of at least 40 bits. Compared with traditional methods for transmitting short words, the present disclosure provides a significant data rate gain. However, the methods included in the present disclosure are directly applicable to transmitting words of any length.

[0099] Examples of the present disclosure employ a first communication device 100 that interacts with a second communication device 300 in a communication system 500, where the first communication device 100 is also referred to herein as a transmitter (TX), and the second communication device 300 is also referred to herein as a receiver (RX).

[0100] Figure 1 A first communication device 100 provided by an example of the present disclosure is shown. In Figure 1 the example shown, the first communication device 100 includes a processor 102, a transceiver 104, and a memory 106. The processor 102 may be coupled to the transceiver 104 and the memory 106 via a communication component 108 known in the art. The first communication device 100 may be used to perform wireless communication and / or wired communication in a wireless communication system and a wired communication system, respectively. The wireless communication capability may be provided by an antenna or an antenna array 110 coupled to the transceiver 104, while the wired communication capability may be provided by a wired communication interface 112 coupled to the transceiver 104. In the present disclosure, the first communication device 100 may be used to perform some actions, which can be understood as the first communication device 100 includes suitable components for performing the actions, such as the processor 102 and the transceiver 104.

[0101] The processor 102 in the first communication device 100 may be referred to as one or more general-purpose central processing units (CPUs), one or more digital signal processors (DSPs), one or more application-specific integrated circuits (ASICs), one or more field programmable gate arrays (FPGAs), one or more programmable logic devices, one or more discrete gates, one or more transistor logic devices, one or more discrete hardware components, and one or more chipsets. The memory 106 in the first communication device 100 may be a read-only memory, a random access memory, or a non-volatile random access memory (NVRAM). The transceiver 104 in the first communication device 100 may be a transceiver circuit, a power controller, an antenna, or an interface for communicating with other modules or devices. In an example, the transceiver 104 in the first communication device 100 may be a separate chipset or may be integrated with the processor 102 in a chipset. In some examples, the processor 102, the transceiver 104, and the memory 106 in the first communication device 100 are integrated in a chipset.

[0102] According to an example of the present disclosure, the first communication device 100 is used to transmit a first character x to the second communication device 300 in the first channel 510. 0 . The first word x 0 is generated based on the information word m. The first communication device 100 is also used for:

[0103] i) receiving a second word p from a second communication device (300) in a second channel 520 i , where the second character p i Indicates the first word x transmitted in the first channel 510 i A group of bits associated and a corresponding set of reliability values Where i is the time index;

[0104] ii) Based on the first word x transmitted i and a set of bits Determine the error word i ;

[0105] iii) Based on a set of reliability values Compression error word i , to generate the first word x i+1 ;as well as

[0106] iv) Transmit the first word x to the second communication device 300 in the first channel 510 i+1 .

[0107] Figure 2 is a flowchart showing a corresponding method 200 that can be executed in the first communication device 100 (e.g., Figure 1 the first communication device 100 shown). The method 200 includes:

[0108] Transmit 202 the first word x to the second communication device 300 in the first channel 510 0 , where the first word x 0 is generated based on the information word m;

[0109] Receive 204 the second word p from the second communication device 300 in the second channel 520 i , where the second word p i indicates a set of bits i associated with the first word x transmitted in the first channel 510 and a corresponding set of reliability values where i is a time index;

[0110] Based on the transmitted first word x i and a set of bits Determine 206 the error word e i ;

[0111] Based on a set of reliability values Compress 208 the error word e i , to generate the first word x i+1 ; and

[0112] Transmit 210 the first word x to the second communication device 300 in the first channel 510 i+1 .

[0113] Figure 3 shows the second communication device 300 provided by an example of the present disclosure. In Figure 3In the example shown, the second communication device 300 includes a processor 302, a transceiver 304, and a memory 306. The processor 302 is coupled to the transceiver 304 and the memory 306 through communication components 308 known in the art. The second communication device 300 can be used for wireless communication and wired communication in a wireless communication system and a wired communication system, respectively. The wireless communication capability can be provided by an antenna or an antenna array 310 coupled to the transceiver 304, while the wired communication capability can be provided by a wired communication interface 312 coupled to the transceiver 304. In the present disclosure, it can be understood that the second communication device 300 can be used to perform some actions, that is, the second communication device 300 includes suitable components for performing the actions, such as the processor 302 and the transceiver 304.

[0114] The processor 302 in the second communication device 300 can be referred to as one or more general-purpose CPUs, one or more DSPs, one or more ASICs, one or more FPGAs, one or more programmable logic devices, one or more discrete gates, one or more transistor logic devices, one or more discrete hardware components, and one or more chip sets. The memory 306 in the second communication device 300 can be a read-only memory, a random access memory, or an NVRAM. The transceiver 304 in the second communication device 300 can be a transceiver circuit, a power controller, an antenna, or an interface for communicating with other modules or devices. In the example, the transceiver 304 in the second communication device 300 can be a separate chip set or integrated with the processor 302 in a chip set. In some examples, the processor 302, the transceiver 304, and the memory 306 in the second communication device 300 are integrated in a chip set.

[0115] According to an example of the present disclosure, the second communication device 300 is used to receive a first word x from the first communication device 100 in a first channel 510 0 , where the first word x 0 is generated based on an information word m. The second communication device 300 is further used to:

[0116] i) Determine a set of bits i associated with receiving the first word x in the first channel 510 and a set of corresponding reliability values where i is a time index;

[0117] ii) Transmit a second word p to the first communication device 100 in a second channel 520 i , where the second word p i indicates a set of bits and a set of corresponding reliability values

[0118] iii) Receive a first word x from a first communication device 100 in a first channel 510 i+1 .

[0119] Figure 4 shows a flowchart of a corresponding method 400 that may be performed in a second communication device 300 (such as Figure 3 shown).

[0120] Method 400 includes:

[0121] Receive 402 a first word x from a first communication device 100 in a first channel 510 0 , wherein the first word x 0 is generated based on an information word m;

[0122] Determine 404 a set of bits i associated with receiving the first word x in the first channel 510 and a set of corresponding reliability values where i is a time index;

[0123] Transmit 406 a second word p to the first communication device 100 in a second channel 520 i , wherein the second word p i indicates a set of bits and a set of corresponding reliability values

[0124] Receive 408 a first word x from a first communication device 100 in the first channel 510 i+1 .

[0125] In the examples of the present disclosure, the technical solutions disclosed herein may be based on interactive channel coding theory. Point-to-point interactive channel coding is a method of transmission over a channel with feedback based on an iterative refinement paradigm. Iterative refinement includes transmitting an uncoded word / message in an initial transmission, and then iteratively refining the information of the transmitted word that the receiver has through subsequent transmissions until the receiver can generate the transmitted word. Each subsequent transmission is based on the word and the feedback information obtained from the receiver through the feedback channel. Iterative refinement is a candidate most suitable for ultra-high reliability transmission because the error probability of these schemes decays doubly exponentially with the codeword length, i.e., significantly faster than the single exponential decay that may be achieved without feedback.

[0126] Multiple forward-feedback iterations involving the first communication device 100 and the second communication device 300 may be performed, where each iteration includes a forward transmission, i.e., transmitting the first word x in the first channel 510 i , and then a feedback transmission, i.e., the second word p in the second channel 520 iTransmission. In each iteration, the first communication device 100 calculates and transmits an accumulated correction for a previous transmission based on information fed back by the second communication device 300 via the second channel 520 in a previous iteration. For example, the iteration can continue until the second communication device 300 can recover the transmitted word without errors and / or until a maximum number of iterations is reached, which will be described more thoroughly in the following disclosure.

[0127] Figure 5 An example of the present disclosure is shown in more detail. In Figure 5 the example of the present disclosure in a set of bits and a set of corresponding reliability values i are determined based on a set of log-likelihood ratio (LLR) z More specifically, each bit in this set of bits i,n can be determined based on the sign of the corresponding LLR z and each reliability value in this set of reliability values i,n can be determined based on the absolute value of the corresponding LLR z.

[0128] Referring to Figure 5 , the first communication device 100 may include a switch 126 that, depending on its switch state, couples its input to an information word input or the output of an error word calculation block 134 at its input. In an initial transmission state, the input of the switch 126 is coupled to the information word input such that the information word m is transmitted to the input of a modulator 130 and the input of a buffer 132 in the first communication device 100, where the buffer 132 stores the information word m for comparison with a first set of corresponding bits . In a retransmission state, the input of the switch 126 is changed to be coupled to the output of the error word calculation block 134.

[0129] The modulator 130 is used to modulate a word and transmit the modulated word to the second communication device 300 via the first channel 510. The output of the buffer 132 including the information word m is coupled to the first input of the error word calculation block 134. The error word calculation block 134 further includes a reliability detection block 136, a compressor 138, and a bit detection block 140. The first communication device 100 also includes a feedback (FB) receiver 124, which is used to receive feedback from the second communication device 300 via the second channel 520. The output of the FB channel receiver 124 is coupled to the reliability detection block 136 and the bit detection block 140 through the second input and the third input of the error word calculation block 134, respectively. The outputs of the reliability detection block 136 and the bit detection block 140 are coupled to the first input and the second input of the compressor 138, respectively. The bit detection block 140 is coupled to the second input of the compressor 138 through a modulo-2 adder, and the modulo-2 adder is also coupled to the output of the buffer 132. The output of the compressor 138 is coupled to the switch 126, which means the output of the compressor 138 becomes the output of the error word calculation block 130.

[0130] Further referring to Figure 5 , the second communication device 300 may include an LLR calculation block 330, and the output of the LLR calculation block 330 is coupled to the input of the quantizer 332. The output of the quantizer 332 is further coupled to the input of the buffer 334 and the input of the feedback (FB) transmitter 324, and the feedback transmitter 324 is used to transmit feedback to the first communication device 100 via the second channel 520.

[0131] In Figure 5 the initial transmission (i.e., Init.TX), the obtained information word m is modulated in the modulator 130 and transmitted by the first communication device 100 to the second communication device 300 in the first channel 510. Based on the corresponding received signal, the second communication device 300 calculates the LLR in the LLR calculation block 330 and quantizes the LLR in the quantizer 332 to obtain a quantized LLR (QLLR) word z (in the form of a vector, etc.). 0 . Then, the QLLR word z 0 is sent back to the first communication device 100 using the FB transmitter 324. In addition, the QLLR word z 0 is stored in the buffer 334. The first communication device 100 obtains the QLLR word z 0 at the FB receiver block 124.

[0132] In Figure 5 the retransmission (i.e., Re.TX), the first communication device 100 is based on the received quantized LLR vector z i-1 and the previously transmitted error word x stored in the buffer 132i-1 An error word x is generated in the error word calculation block 134 i . The previously transmitted error word x i-1 and a set of bits indicated by the quantized LLR vector z i-1 are subjected to element-wise modulo-2 addition to obtain a word e compressed in the compressor 138 , thereby generating the error word x i . Then, the first communication device 100 modulates the error word x i and transmits it to the second communication device 300 over the first communication channel 510. Based on the corresponding received signal, the second communication device 300 calculates the LLR in the LLR calculation block 330 and quantizes the LLR in the quantizer 332 to obtain the quantized LLR word z i . The quantized LLR word z is sent back to the first communication device 100 using the FB transmitter 324 i . In addition, the quantized LLR word z i is stored in the buffer 334, so each quantized LLR word z i is stored in the buffer 334 for later decoding. i

[0133] Figure 6 More details are shown for encoding the information word m when using several coding iteration steps. Each iteration step includes a forward transmission from the first communication device 100 to the second communication device 300, and then includes a subsequent feedback transmission from the second communication device 300 to the first communication device 100. In Figure 6 , the solid arrows represent the transfer / transmission of data, and the dashed arrows represent the transfer / transmission of control information.

[0134] Before the initial transmission, the communication device scheduling the transmission (i.e., the first communication device 100 or the second communication device 300, depending on where the scheduler is running) selects a modulation scheme based on the channel quality assessment of the first channel 510 and further determines a set of time-frequency channel resources. Then, the communication device sends information about the selected modulation scheme and time-frequency resources in a given forward channel time-frequency grid for transmission to the other communication device. This can be performed, for example, through a control channel such as the Physical Downlink Control Channel (PDCCH) of 3GPP NR, where a suitable control message including a time-frequency resource indication is sent.

[0135] Thereafter, referring to Figure 6 , in the first iteration step, i.e., when the time index i = 0:

[0136] – The first communication device 100 modulates the obtained information word m in the modulator 130 to obtain the modulated first word x 0 ;

[0137] – The first communication device 100 uses the previously determined time-frequency resources for transmission and transmits the modulated first word x in the communication signal to the second communication device 300 through the first channel 510 (i.e., the forward channel) 0 ;

[0138] – The second communication device 300 receives the first word x affected by the first channel 510 0 , that is, y 0 ;

[0139] – The second communication device 300 calculates the first set of QLLR z of the received first word y in the LLR calculation block 330 and the quantizer 332 based on the communication signal received from the first communication device 100 0 . The first set of QLLR z 0 can be stored in the local buffer 334 of the second communication device 300, as 0 shown; and Figure 6 ;

[0140] – The second communication device 300 also sends the first set of QLLR z to the first communication device 100 through the second channel 520 (i.e., the feedback channel) 0 as feedback

[0141] Further referring to Figure 6 , in the second iteration step, that is, when the time index i = 1:

[0142] – The first communication device 100 determines the first error word e in the error word calculation block 134 based on the information word m and the first set of QLLR

[0143] z 0 received through the second channel 520 in the previous iteration 1 ;

[0144] – The first communication device 100 further compresses the first error word e in the compressor 138 of the error word calculation block 134 i to generate the first word x 0 ;

[0145] – The first communication device 100 modulates the first word x 1 in the modulator 130

[0146] – The first communication device 100 transmits the modulated first word x to the second communication device 300 in the communication signal through the first channel 510 1 ;

[0147] – The second communication device 300 receives a first word x affected by the first channel 1 , which is y 1 ;

[0148] – The second communication device 300 calculates a second set of QLLRs z of the received first error word y 1 based on the received communication signal 1 ,

[0149] and stores the second set of QLLRs z 1 in buffer 334;

[0150] – The second communication device 300 also sends the second set of QLLRs z 1 to the first communication device 100 via the second channel 520

[0151] as feedback. Figure 6 With further reference to

[0152] in the i-th (where i = 1,…, D) iteration step: i – The first communication device 100 determines a first error word e in the error word calculation block 134 i based on the information word m and the i-th set of QLLRs z received via the second channel 520 in the previous iteration step

[0153] – The first communication device 100 further compresses the first error word e in the compressor 138 of the error word calculation block 134 i to generate a first word x i ;

[0154] – The first communication device 100 modulates the first word x i in the modulator 130

[0155] – The first communication device 100 transmits the modulated first word x i to the second communication device 300 in the communication signal via the first channel 510

[0156] – The second communication device 300 receives a first word x affected by the first channel i , which is y i ;

[0157] – The second communication device 300 calculates the i-th set of QLLRs z of the received first error word y i based on the received communication signal i ,

[0158] and stores the i-th set of QLLRs in buffer 334;

[0159] – The second communication device 300 also sends the i-th set of QLLRz to the first communication device 100 via the second channel 520 i as feedback.

[0160] The above process is repeated D times of iteration, where the parameter D depends on at least one stopping condition. Therefore, in the example of the present disclosure, the first communication device 100 is used to repeat its steps i) to iv) D times, where i = 0, …, D−1, and D depends on at least one stopping condition. Correspondingly, the second communication device 300 is also used to repeat its steps i) to iv) D times to obtain multiple sets of bits and multiple sets of corresponding reliability values

[0161] In the example of the present disclosure, the first stopping condition may be the error-free reception of the first word x D in the second communication device 300. Therefore, the iteration continues until the first communication device 100 determines that no error has occurred in the forward transmission by comparing a set of bits and the first word x i-1 , that is, until no error has occurred in the forward transmission within a certain number of iterations D. When the first communication device 100 determines that no error has occurred in a given forward transmission, the first communication device 100 transmits a first control signal 530 to the second communication device 300, as Figure 6 shown. The first control signal 530 indicates the error-free reception of the first word x D , for example, a termination signal in the form of an acknowledgement (ACK) to stop the coding process.

[0162] In the example of the present disclosure, the second stopping condition may be the latency requirement of the information word m. This may indicate that the parameter D may be a design parameter. For example, this design parameter may be given by a standard defined by 3GPP, etc. For example, in NR, the parameter D may have a first value corresponding to the Ultra Reliable Low Latency Communication (URLLC) service and a second different value corresponding to the enhanced Mobile Broad Band (eMBB) service, because these two services have different latency requirements. In the example, since the latency requirement of the URLLC service is more stringent, the first value is less than the second value.

[0163] When D times of iteration have been performed, in the example of the present disclosure, the second communication device 300 uses its decoder 322 to start the decoding process, as Figure 6 shown.

[0164] Figure 7More specifically, a decoder 322 in the second communication device 300 provided by the present disclosure is shown. The decoder 322 includes a bit detector block 350 and a plurality of error correction blocks 352. The bit detector block 350 is coupled to the error correction blocks 352, and the plurality of error correction blocks 352 are coupled in series with each other.

[0165] When a termination signal (such as an ACK or NACK signal) is received through the control channel, the decoder 322 may declare a decoding failure (for example, in the case of NACK) or start a decoding operation (for example, in the case of ACK). The bit detector block 350 converts the QLLR into bits to obtain an error word x D . Based on the error word x D , the error correction block 352 performs error correction on a set of bits indicated in the QLLR word z D-1 to obtain a corresponding error word x D-1 . The above steps are repeated until the error word x 1 is obtained. Finally, the error word x 1 is used in the first error correction block 352 to obtain the decoded message m.

[0166] In other words, the second communication device 300 decodes multiple sets of bits and multiple sets of corresponding reliability values to obtain the information word m. In the example of the present disclosure shown in Figure 7 , decoding multiple sets of bits and multiple sets of corresponding reliability values may include: i) correcting errors in a set of bits i based on the first word x and a set of corresponding reliability values to obtain the first word x i-1 ; repeating i) D times to obtain the first word x 0 . Finally, the information message m is determined based on the first word x 0 .

[0167] However, if the maximum number of iterations D has been reached and the first communication device 100 determines that there are still remaining errors in the forward transmission to the second communication device 300, the first communication device 100 may send a control signal indicating a negative acknowledgement (NACK) to the second communication device 300, as also shown in Figure 6 . Thereafter, the first communication device 100 terminates the transmission of the first word x i to the second communication device 300.。When the second communication device 300 receives a control signal indicating NACK, it can output a decoding failure message. Therefore, the aforementioned first stop condition and second stop condition can be combined. After decoding failure, the first communication device 100 can attempt to transmit the same information word for the second time, or discard the word without performing other attempts.

[0168] According to the above iterative coding process, the received codeword (as Figure 6 shown) can be a combination of multiple first words, generated by subsequent transmissions of the compressed error words (i.e., the first words). Therefore, the codeword length is variable because the number of coding iterations and the length of each first word are not known a priori.

[0169] As described above, in the example of the present disclosure, based on the aforementioned set of log-likelihood ratios z i to determine a set of bits and a corresponding set of reliability values The LLR can be calculated in the i-th coding iteration in the following manner: Calculate the LLR of a given transmitted bit according to the conventional formula:

[0170]

[0171] where y i is the received signal in the (i + 1)-th iteration, and N i is its length.

[0172] The quantization of each LLR is as follows:

[0173]

[0174] Here, is the reliability level calculated based on , and sign(a) is a function that takes the value +1 if a > 0 and -1 otherwise.

[0175] Figure 8a The error word calculation block 134 in the first communication device 100 is shown in more detail, while Figure 8b the error correction block 352 in the second communication device 300 provided by the example of the present disclosure is shown in more detail. The compression step in the first communication device 100 and the expansion step in the second communication device 300 are for reducing the channel resource utilization rate, thereby improving the spectral efficiency. It should be noted that x 0 ≡m.

[0176] Refer to Figure 8aThe compressor 138 of the error word calculation block 134 in the first communication device 100 may also include a demultiplexer (DEMUX) 144 and a multiplexer (MUX) 142. The output of the MUX 142 is coupled to the switch 126. The error word calculation block 134 also includes a modulo-2 adder 146 (also in Figure 5 ), a modulo-2 adder 146 is coupled between the DEMUX 144 and the bit detection block 140 and the buffer 132.

[0177] refer to Figure 8b Each error correction block 352 in the second communication device 300 may include an expander block 354, which in turn includes a DEMUX 344 and a MUX 342. A modulo-2 adder 346 is also coupled between the MUX 342 and the bit detection block 140 and the buffer 334.

[0178] refer to Figure 8a In the first communication device 100, the i-th first word x is calculated i This can be done based on the following steps:

[0179] - Determine the set of QLLRz received i-1 A group of bits indicating in,

[0180]

[0181] -Based on the first word x i-1 Detect a group of bits The error in getting the word

[0182]

[0183] in, It represents the element-wise sum modulo 2.

[0184] - Determine the set of QLLRz received i-1 The corresponding reliability value of the indicated set in,

[0185]

[0186] -According to DEMUX 144 Word Demultiplexing is performed to obtain L subwords

[0187] l=1,…,L, where the lth subword is defined as

[0188]

[0189] - Compress each sub - word in the compressor 138 to obtain a new compressed error sub - word

[0190] -(Optional step) If the new compressed error sub - word is longer than the new sub - word then use the word to replace and

[0191] - Multiplex the new compressed error sub - word in the MUX 142 to obtain a new error word x forwarded to the switch 126 i .

[0192] When the new error sub - word generated by the compressor 138 is longer than the new sub - word the optional step provides the further advantage of using fewer channel resources.

[0193] The compression of the new error sub - word can be accomplished by the compressor 138 using any conventional data compression algorithm. The compression algorithm is the same for all reliability levels. However, the compression level must match the bit error rate in each level. For example, close - to - optimal compression can be achieved by Huffman coding using a dictionary adapted to the bit error rate in each reliability level.

[0194] Refer to Figure 8b Calculating the (i - 1)th first word x in the second communication device 300 i-1 can be performed according to the following steps:

[0195] - Determine a set of bits i-1 indicated by a set of QLLRz where

[0196]

[0197] - Determine a corresponding set of reliability values i-1 indicated by a set of QLLRz where

[0198]

[0199] - Demultiplex the ith error word x i in the DEMUX 344 to obtain the compressed error sub - word

[0200] -(Optional step) For the compressed error sub - words that have been sub - word replaced in the first communication device 100 skip the next step;

[0201] - Expand each compressed error sub - word in the expander block 354 to obtain sub - words

[0202]

[0203] - Multiplex the sub - words in the MUX 342 according to a set of corresponding reliability values to obtain the word e i ; and

[0204] - Based on x i correct the errors in a set of bits in the modulo - 2 adder 346 to obtain the following first word x i-1 :

[0205]

[0206] When the error sub - words generated by the compressor are longer than the sub - words the optional step provides the further advantage of using fewer channel resources.

[0207] Figure 9 Shows an example of the present disclosure when the feedback signal from the second communication device 300 to the first communication device 100 is compressed before being transmitted in the second channel 520. In Figure 9 the solid arrows represent the transfer of data and the dashed arrows represent the transfer of control signals.

[0208] The error word calculation block 134 calculates the first word x i-1 based on the word x i-1 and z i . The first word x i is transmitted to the second communication device 300 through the first channel 510, and the second communication device 300 obtains the received signal y i . Based on the received signal y i the second communication device 300 calculates and quantizes the LLR in the LLR calculation block 330 and the quantizer 332 (shown here as separate blocks) to obtain the QLLR word z i . Based on the QLLR word z i the generator matrix selection block 360 selects the generator matrix G i from the generator matrix library.

[0209] Using the system encoding block 362, based on the QLLR word z i and the selected generator matrix G i generate the second word p i The second communication device 300 selects the generator matrix G in the matrix selection block 360 based on the length of the first word x i Thereafter, by encoding a set of bits i in the encoding block 362 a corresponding set of reliability values and the selected generator matrix G i the second word p is obtained i Finally, the second communication device 300 transmits the second control signal 540 to the first communication device 300 via the control channel. The second control signal 540 indicates the length of the second word p i

[0210] More specifically, the QLLR vector is encoded to obtain the parity check bit p i vector based on QLLR z i For example, the second communication device 300 generates the codeword Θ i = G i z i where G i is the generator matrix of the systematic error correcting code; since the code is a systematic code, the form of the codeword is Θ i = [z i ; p i , that is, the initial N i symbols of Θ i are the same as the encoder input word z i and the remaining bits are parity check bits. The second communication device 300 transmits the parity check bit p i back to the first communication device 100 via the second channel 520

[0211] The first communication device 100 receives the second control signal 540 from the second communication device 300 and extracts the information about the length of the second word p i from the second control signal 540 in the decoding block 160. Based on the length of the second word p i the first communication device 100 determines the generator matrix G i used by the second communication device 300. Finally, the first communication device 100 obtains a set of bits i and a corresponding set of reliability values i based on the generator matrix G and the second word p ​For example, the systematic generator matrix of any linear binary block code such as a Low Density Parity Check (LDPC) code, a Bose-Chaudhuri-Hocquengham (BCH) code, an extended BCH code, or a systematic polar code can be used as G i .

[0212] The first communication device 100 receives a corrupted parity bit word wherein, represents the error introduced by the second channel 520. The first communication device 100 forms a corrupted codeword based on x i and q i For example, when the code is a systematic code, the corrupted codeword is Then, the first communication device 100 decodes i using the generator matrix G to obtain a decoded codeword If the error correction capability t of the code specified by G i is at least equal to the total number of errors occurring in the forward transmission and the feedback transmission, i.e., the number of errors in ε i plus the number of errors in i , then the decoding block 160 generates a decoded codeword Therefore, the first communication device 100 determines z from , and then calculates the compressed error word x i , which is transmitted to the second communication device 300 in the first channel 510 in the last step. i+1

[0213]

[0214] One advantage of this example is that when p i is shorter than z i , i.e., when this example achieves feedback compression through coding, transmitting p i reduces the data rate required on the second channel 520. Another advantage of this example is that, unlike uncoded feedback, even when errors occur in the feedback transmission, the first communication device 100 can determine the forward channel error vector ε i .

[0214] In addition, according to the example of the present disclosure, when in the initial transmission, |λ 0,k | < ε (k = 1, …, K), the output value of the quantizer 332 is z 0,k = 0; when in the i-th retransmission, |λ i,n | < ε (n = 1, …, N i ), the output value of the quantizer 332 is z i,n ​= 0. Here, ε is an arbitrary non - negative real value. The bits received with a reliability of 0 are erased bits. The erased bits are re - transmitted by the encoder in the next iteration without any compression. Thus, in the example of the present disclosure, the first communication device 100 is used to generate the first word x in the following manner i+1 : Add bits to the first word x transmitted i to generate the first word x i+1 , where the reliability value of the first word transmitted is 0.

[0215] The advantage of the above example is that it avoids compressing the least reliable bits, i.e., the bits with a reliability level of 0. The bit error rate of the least reliable bits is very high, close to 0.5. Therefore, compression provides negligible theoretical gain and even shows a slight deterioration in actual implementation.

[0216] In other examples of the present disclosure, the first communication device 100 is used to add at least one redundant bit so that the number of bits of the first word x i+1 adapts to the frequency and time resources available for transmission. Generally, the number of bits H i that can be accommodated in the time - frequency resources in the i - th re - transmission is greater than the number of bits N i in the error word e i (H i > N i ). In this case, the error word e i can be encoded by adding H i −N i redundant bits, thereby generating the i - th encoded error word g i with a length of H i bits. The i - th encoded error word g i is modulated to obtain the transmitted signal x i , and then transmitted to obtain the corresponding received signal y i in the second communication device 300. The received signal is decoded by the second communication device 300 to obtain the decoded soft error word h i . Then, the decoded soft error word h i can be sent to the input of the quantizer 332.

[0217] The above example is applicable to LTE and / or NR transmissions, etc., where in each transmission, the number of time - frequency channel resources is frequency - allocated in resource block granularity.

[0218] In other examples of the present disclosure with reference to Figure 10 , the modulation used on the first channel 510 can be BPSK modulation or QPSK modulation. Thus, the modulator 130 can be Figure 10The BPSK / QPSK modulator shown. When the modulation is BPSK modulation, the signal generated by the modulator 130 can be 1 - 2m in the first encoder iteration and 1 - 2e in subsequent encoder iterations i , i = 1, …, D. A set of bits and a set of corresponding reliability values are indicated by the quantized word z i =(y i,1 , …, y i,K ) obtained by quantizing the received signal y i =(z i,1 , …, z i,K ).

[0219] Quantization can be performed according to a conventional quantization method as follows:

[0220]

[0221] Here, τ=(τ 0 = 0, τ 1 , τ 2 , …, τ L-1 , τ L = ∞) is a vector of quantization boundary values such that for any l = 1, …, L, τ l-1 ≤τ l . Quantization of the zero reliability (erasure) level, as described in the third example above, is achieved by setting τ 0 >0.

[0222] When the modulation is QPSK modulation, the I / Q separator 370 in the second communication device 300 converts the received signal y i into a pair of real signals y i,I and y i,Q , which are converted from parallel to serial in the parallel / serial block 372. Then, the real signals are quantized in the quantizer 332 to obtain a signal z indicating a set of bits and a set of corresponding reliability values i .

[0223] Quantization of the memoryless channel output results in an equivalent Discrete Memoryless Channel (DMC), whose capacity is determined based on its transition probability matrix. In turn, the transition probability matrix is determined by the quantization boundary values τ. Therefore, for any given number of reliability levels L, it is possible to find the optimal quantization boundary vector τ *. In this case, LLR calculation is not required. The above example of using BPSK / QPSK modulation in the first channel 510 has the following advantages: Simplifying the system when using low-order modulation, thus achieving the goal of providing a very low error rate and high spectral efficiency while having limited complexity.

[0224] Performance Evaluation

[0225] Evaluate the performance gain provided by the examples of the present disclosure relative to traditional codes based on spectral efficiency (SE). The spectral efficiency provided by the examples of the present disclosure is defined as

[0226]

[0227] where E[·] represents expectation and N is the codeword length.

[0228] Analytically evaluate the SE achievable by the above technical solutions and give a delay analysis through simulation. The results show that the examples of the present disclosure meet the delay constraints of machine control applications. In all evaluations, it is assumed that the first example uses compressed error words for transmission, the second example uses a zero reliability level, and the third example uses BPSK / QPSK modulation.

[0229] With BPSK modulation, the channel input is v 0,k = 1 - 2m k , k = 0, …, K - 1; the corresponding channel output is y 0,k = v 0,k + w 0,k , where w 0,k is an iid Gaussian RV with zero mean and variance 1 / ρ, where ρ is the Signal-to-Noise Ratio (SNR). The quantizer output is given by equation (10).

[0230] Given the channel input, the probability of the quantizer output is given by:

[0231] p l = P{z 0,k = +l|m k = 0} = P{z 0,k = -l|m k = 1}, (12)

[0232]

[0233] The probability that each element in the quantizer output z 0 has reliability l is given by:

[0234]

[0235] Therefore, z 0 follows a multinomial distribution with probability q = (q 1 , …, q L ).

[0236] With the above definition, the elements of in Equation (6) are non-zero with probability p l / q l . Therefore, efficient data compression of 1,l results in a compressed sub-word of length N where

[0237]

[0238] Here, H 2 (x) = -x log 2 x - (1 - x) log 2 (1 - x). The lower bound in Equation (15) can be achieved in reality with a small gap even when is very short. Therefore, by summing the lower bounds over all sub-words , a strict lower bound on the length of the first retransmission is obtained as follows:

[0239]

[0240] The codeword is c = (m, x 1 , …, x D ), where D is the number of retransmissions, i.e., a random variable that can take any non-negative integer value in the set . To estimate the achievable SE, the average codeword length needs to be determined as

[0241]

[0242] where it is assumed that the number of retransmissions can reach infinity. Here, for ease of notation, N 0 ≡ K.

[0243] Assume that the lower bound in Equation (15) has been reached. Therefore, given the length of the (i - 1)-th retransmission N i-1 , the expected length of the i-th retransmission N i-1 is given by:

[0244]

[0245] where n = (n 1 , …, n L ), and n l is The length. The inner summation in equation (19) is the expected number of times of observing the reliability level l over the N i-1 elements in the (i - 1)-th retransmission, which is a known property of the multinomial distribution:

[0246]

[0247] Combining equations (19) and (20) gives

[0248]

[0249] where p = (p 1 , …, p L ), and the L-level entropy H L is defined as:

[0250]

[0251] Finally, E(N i ) can be obtained from equation (21) as follows:

[0252]

[0253] where is the (unknown) probability density function of N i-1 . Combining equations (23) and (17) gives

[0254]

[0255] Based on equation (24), the minimization of the average codeword length can be achieved by finding the vectors p, q that minimize H L . On the BI-AWGN channel with SNR ρ, p and q can be calculated as follows:

[0256]

[0257] where τ = (τ 0 = 0, τ 1 , τ 2 , …, τ L-1 , τ L = ∞) is the vector of quantization boundary values such that for any l = 1, …, L, τ l-1 ≤ τ l . Clearly, given SNR ρ, τ determines p, q. Therefore, H L (p, q) can be rewritten as H L (τ, ρ): It can be seen that for a given number of levels L and SNR ρ, minimizing E(N) requires finding the quantizer τ that minimizes H L (τ, ρ)* , as follows:

[0258]

[0259] An algorithm is implemented to find the optimal quantizer that maximizes the mutual information between the channel input and the quantizer output. Finally, due to data processing inequality, there exists

[0260]

[0261] where C BI-AWGN (ρ) is the symmetric capacity of a BI-AWGN channel with SNR ρ. By combining Equation (27) and (24), it can be derived from Equation (11) that the examples of the present disclosure can achieve any SE that is arbitrarily close to the BI-AWGN channel capacity:

[0262]

[0263] Although Equation (28) indicates that approaching the BI-AWGN capacity with an arbitrarily small gap requires L to grow to infinity, which shows that even with a small number of reliability levels (e.g., L = 1, 2), the AIC SE is very close to C BI-AWGN , thus providing a significant improvement over conventional technical solutions with lower complexity.

[0264] The achievable SE provided by the examples of the present disclosure is evaluated through Monte Carlo simulations and compared with the SE achieved by conventional transmissions and corresponding performance bounds. The exact performance bound of conventional transmissions is called the normal approximation of the achievability bound. This bound provides the achievable SE for finite codeword length transmissions over binary input AWGN (BI-AWGN) without feedback. The state-of-the-art codes, namely NR polar codes and eBCH codes, have performance very close to the PPV bound.

[0265] Performance evaluation is carried out through Huffman coding for error word compression. The data compressor in the first communication device 100 uses a dictionary with a fixed size S = 2 s , where each dictionary is generated according to the bit error probability p l corresponding to the reliability level. When L = 1, s = 5 is sufficient to achieve almost optimal performance when the SNR is below 2 dB, while s = 12 is required when the SNR is above 5 dB. The quantization boundaries (τ 0 , τ 1 , …, τ L = ∞) are selected to maximize the capacity of the equivalent DMC. The number of retransmissions is not restricted.

[0266] Figure 11It is shown that the examples of the present disclosure can provide ultra-high reliability transmission, where the SE is greater than 90% of the BI-AWGN channel capacity even when the SNR is below 0 dB; for the same SNR, the SE of traditional codes is less than 50% of the QPSK channel capacity. The examples of the present disclosure are superior to deep codes (traditional technical solutions based on deep learning), with an SNR gain of approximately 1.8 dB. The performance of the examples of the present disclosure is slightly better than that of variable length feedback convolutional codes (VLF-CC) (traditional technical solutions based on variable length convolutional coding using feedback), although the complexity of VLC-CC is much higher. In Figure 11 it, the x-axis represents the SNR in dB, and the y-axis represents the SE in bits / s / Hz.

[0267] In NR time division duplex (TDD), transmissions are organized in time slots consisting of 14 orthogonal frequency division multiplexing (OFDM) symbols. According to a set of predefined patterns in Table 11.1.1-2 of 3GPP TS38.213 V15.5.0, each symbol can be semi-statically configured as a downlink (D) symbol, an uplink (U) symbol, or a flexible (F) symbol. Table 1 outlines Table 11.1.1-2 of 3GPP TS 38.213 V15.5.0, where "D" / "U" symbols can only be used for downlink / uplink transmissions, while "F" symbols can be used for downlink or uplink, as shown by L1 dynamic scheduling. Any guard intervals required for TDD operation (e.g., downlink-uplink switching time, user equipment (UE) timing advance, etc.) are obtained from the flexible symbols. Therefore, at least one "F" symbol serves as the guard time between "D" symbols and "U" symbols.

[0268] Table 1: Time slot format with normal cyclic prefix (Table 11.1.1-2 of 3GPP TS 38.213 V15.5.0).

[0269]

[0270]

[0271] The time slot duration depends on the subcarrier spacing, as shown in Table 2.

[0272] Table 2: Subcarrier spacing and time slot duration, where μ is the NR base parameter (numerology parameter).

[0273] μ Subcarrier Spacing [kHz] Time Slot Duration 0 15 1ms 1 30 500μs 2 60 250μs 3 120 125μs 4 240 62.5μs

[0274] The first option considers TDD time slot format 2 in Table 1. Format 2 is fully flexible, meaning that transmissions can be dynamically scheduled according to any downlink - uplink pattern. For example, the pattern in Table 3 provides three complete iterations plus the initial part of the 4th iteration to be completed in the next time slot. At least one guard symbol between D / U and U / D is required to accommodate the timing advance in UE and UE / gNB processing times.

[0275] Table 3: Downlink - uplink patterns for AIC based on TDD time slot format 2, where "D" = downlink, "U" = uplink, "G" = guard ( = no transmission).

[0276]

[0277] The second option considers TDD time slot format 50 in Table 1. Format 50 can be used to provide two complete Physical (PHY) layer iterations in each time slot, as shown in Table 4.

[0278] Table 4: Downlink - uplink patterns for AIC based on TDD time slot format 50, "D" = downlink, "U" = uplink, "G" = guard ( = no transmission).

[0279]

[0280] For a given air - interface latency T AI , the maximum number of PHY layer iterations achieved by NR TDD for each basic parameter μ and TDD time slot format is obtained by combining the time slot timing information in Table 2 with the patterns in Tables 3 and 4. The result for T AI = 500 μs is shown in Table 5.

[0281] Table 5: Maximum number of PHY layer iterations, where T AI = 500 μs.

[0282]

[0283]

[0284] The above analysis shows that there is a time budget of up to 28 iterations (see Table 5), where the air - interface latency T AI = 500 μs.

[0285] Compared with traditional Long-Term Evolution / New Radio (LTE / NR) Hybrid Automatic Repeat Request (HARQ), the examples of the present disclosure perform more retransmissions. However, typical HARQ retransmissions have the same or similar length as the initial transmission, while the examples of the present disclosure shorten the length of the retransmissions. This may make the examples of the present disclosure more efficient in terms of channel resource utilization. In addition, each HARQ retransmission is completed after a failure of the Forward Error Correction (FEC) decoding attempt in the receiver. FEC decoding consumes energy / time, so any HARQ retransmission cycle may take a long time and / or consume a large amount of energy in the receiver. The retransmissions in the examples of the present disclosure do not require any decoding because they are part of the encoding process. Therefore, the time required for retransmissions is much shorter and the computational workload is also less, that is, decoding is only completed once after all error words have been transmitted. In addition, the proportion of codewords that require 20 or more retransmissions can be ignored.

[0286] In the examples of the present disclosure, the first communication device 100 may be a network access node, and the second communication device 300 may be a client device. In other examples, the first communication device 100 may be a client device, and the second communication device 300 may be a network access node, that is, the opposite case.

[0287] The client devices in this disclosure include, but are not limited to: UEs (such as smart phones, cellular phones, cordless phones, session initiation protocol (SIP) phones, wireless local loop (WLL) stations, personal digital assistants (PDAs)), handheld devices with wireless communication capabilities, computing devices or other processing devices connected to wireless modems, in-vehicle devices, wearable devices, integrated access and backhaul (IAB) nodes (such as mobile cars or devices installed in cars), drones, device-to-device (D2D) devices, wireless cameras, mobile stations, access terminals, subscriber units, wireless communication devices, wireless local access network (WLAN) stations, tablet computers with wireless capabilities, laptop embedded devices, universal serial bus (USB) dongles, wireless customer-premises equipment (CPE), and / or chipsets. In the internet of things (IoT) scenario, the client device can represent a machine or another device or chipset that communicates with another wireless device and / or network device.

[0288] The UE can also be referred to as a mobile phone, cellular phone, computer tablet or laptop with wireless capabilities. In this context, the UE can be, for example, a portable, pocket-sized, handheld, computer-built or in-vehicle mobile device capable of transmitting voice and / or data via a radio access network to another entity (such as another receiver or server). The UE can be a station (STA), i.e., any device that includes an IEEE 802.11-compliant media access control (MAC) and physical layer (PHY) interface connected to a wireless medium (WM). The UE can also be used for communication in 3GPP-related LTE and advanced LTE, in WiMAX and its evolutions, and in fifth-generation wireless technologies (such as NR).

[0289] In addition, the network access nodes in this document can be represented as wireless network access nodes, access network access nodes, access points, or base stations, such as radio base stations (RBS). In some networks, it can be referred to as a transmitter, "gNB", "gNodeB", "eNB", "eNodeB", "NodeB", or "B node", depending on the technology and terminology used. Based on the transmission power and cell size, wireless network access nodes can have different categories, for example, macro eNodeB, home eNodeB, or pico base stations. The wireless second communication device can be a station (STA), that is, any device including an IEEE 802.11-compliant MAC and PHY interface connected to the wireless medium. The wireless second communication device can also be a base station corresponding to a 5G wireless system.

[0290] In addition, any method provided by the examples of the present disclosure can be implemented in a computer program having code components that, when run by a processing component, cause the processing component to execute the steps in the method. The computer program is included in a computer-readable medium of a computer program product. The computer-readable medium can basically include any memory, such as read-only memory (ROM), programmable read-only memory (PROM), erasable PROM (EPROM), flash memory, electrically erasable EPROM (EEPROM), or hard disk drive.

[0291] In addition, those skilled in the art recognize that examples of the first communication device 100 and the second communication device 300 include the necessary communication capabilities in the form of functions, modules, units, elements, etc. for implementing the technical solution. Examples of other such modules, units, elements, and functions are: processors, memories, buffers, control logic, encoders, decoders, rate matchers, de-rate matchers, mapping units, multipliers, decision units, selection units, switches, interleavers, de-interleavers, modulators, modems, inputs, outputs, antennas, amplifiers, receiving units, transmitting units, DSPs, medium specific decoders (MSD), trellis coding modulation (TCM) encoders, TCM decoders, power units, power feeds, communication interfaces, communication protocols, etc., which are appropriately arranged together to execute the technical solution.

[0292] Specifically, one or more processors in the first communication device 100 and the second communication device 300 may include, for example, one or more instances of a central processing unit (CPU), a processing unit, a processing circuit, a processor, an application specific integrated circuit (ASIC), a microprocessor, or other processing logic that can interpret and execute instructions. Thus, the expression "processor" may refer to a processing circuitry that includes multiple processing circuits (such as any, part, or all of those mentioned above). The processing circuitry may also perform data processing functions for input, output, and processing of data, and the data processing functions include data buffering and device control functions, such as call processing control, user interface control, and the like.

[0293] Finally, it should be understood that the present disclosure is not limited to the above examples, but also relates to and includes all examples within the scope of the appended independent claims.

Claims

1. A first communication device (100) for a communication system (500), characterized in that, the first communication device (100) is configured to: Transmit a first word x to a second communication device (300) in a first channel (510) 0 , where the first word x 0 is generated based on an information word m; i) Receive a second word p in the second channel (520) from the second communication device (300) i , wherein the second word p i indicates a set of bits associated with the first word x transmitted in the first channel (510) i and a corresponding set of reliability values wherein i is a time index ​ ii) Based on the first transmitted word x i and the set of bits determine the error word e i , iii) Based on the set of corresponding reliability values Compress the error word e i to generate the first word x i+1 and (iv) Transmit the first word x in the first channel (510) to the second communication device (300) i+1 ; wherein, the first communication device (100) is configured to: repeat i) to iv) D times, where i = 0, …, D-1 and D depends on at least one stop condition.

2. The first communication device (100) according to claim 1, characterized in that, The first stop condition is the first character x D For error-free reception in the second communication device (300), the first communication device (100) is configured to: Transmit a first control signal (530) to the second communication device (300), wherein the first control signal (530) indicates an error-free reception of the first word x D of.

3. The first communication device (100) according to claim 1 or 2, characterized in that, The second stop condition is the latency requirement of the information word m.

4. The first communication device (100) according to claim 1 or 2, characterized in that, The error character e i indicates the first character x i and the set of bits with bits having different values.

5. The first communication device (100) according to claim 1 or 2, characterized in that, The set of bits and the corresponding set of reliability values are determined based on a set of log-likelihood ratios LLR z i respectively.

6. The first communication device (100) according to claim 1 or 2, characterized in that, Generate the first character x i+1 including: At the first word x of the transmission i Add bits to generate the first word x i+1 , wherein the reliability value of the first word x of the transmission i is 0.

7. The first communication device (100) according to claim 1 or 2, characterized in that, the first communication device (100) is configured to: Add at least one redundant bit so that the number of bits of the first word x i+1 adapts to the frequency and time resources for transmission.

8. The first communication device (100) according to claim 1 or 2, characterized in that, the first communication device (100) is configured to: Receiving a second control signal (540) from the second communication device (300), wherein the second control signal (540) indicates the length of the second word p i ; Determine the generator matrix G based on the length of the second word p i ; and i ; and Based on the generating matrix G i and the second word p i , obtain the set of bits and the set of corresponding reliability values 9. A second communication device (300) for a communication system (500), characterized in that, the second communication device (300) is configured to: Receive a first word x from a first communication device (100) in a first channel (510) 0 , wherein the first word x 0 is generated based on an information word m; i) Determine a set of bits i associated with receiving the first word x in the first channel (510) and a corresponding set of reliability values where i is a time index ii) Transmit a second word p to the first communication device (100) in a second channel (520) i , wherein the second word p i indicates the set of bits and the set of corresponding reliability values iii) Receive a first word x from the first communication device (100) in the first channel (510) i+1 ; wherein, the second communication device (300) is configured to: Repeat i) to iii) D times to obtain multiple groups of bits and multiple groups of corresponding reliability values where i = 0, …, D−1 and D depends on at least one stopping condition.

10. The second communication device (300) according to claim 9, characterized in that, The first stop condition is the first character x D Error-free reception in the second communication device (300), the second communication device (300) being adapted to: Receiving a first control signal (530) from the first communication device (100), wherein the first control signal (530) indicates the first word x D Error-free reception in the second communication device (300).

11. The second communication device (300) according to claim 9 or 10, characterized in that, The second stop condition is the latency requirement of the information word m.

12. The second communication device (300) according to claim 9 or 10, characterized in that, The set of bits and the corresponding set of reliability values are determined based on a set of LLR z i determined.

13. The second communication device (300) according to claim 9 or 10, characterized in that, the second communication device (300) is configured to: Decode the multiple groups of bits and the multiple groups of corresponding reliability values to obtain the information word m.

14. The second communication device (300) according to claim 13, characterized in that, For the multiple groups of bits and the multiple groups of corresponding reliability values decoding includes: i) Based on the first word x i and the corresponding set of reliability values correct errors in a set of bits to obtain the first word x i-1 ; Repeat i) D times to obtain the first character x 0 ; and Based on the first character x 0 Determine the information word m.

15. The second communication device (300) according to claim 9 or 10, characterized in that, the second communication device (300) is configured to: Generate the generator matrix G based on the length selection of the first character x i i ;​ Based on the set of bits the corresponding set of reliability values and the generation matrix G i , obtain the second word p i ; and Transmit a second control signal (540) to the first communication device (100), wherein the second control signal (540) indicates the length of the second word p i of.

16. A method (200) for a first communication device (100), characterized in that, the method (200) comprises: Transmit (202) a first word x to a second communication device (300) in a first channel (510) 0 , wherein the first word x 0 is generated based on an information word m; i) receiving (204) a second word p from the second communication device (300) in a second channel (520); i , wherein the second character p i Indicator and the first word x transmitted in the first channel (510) i A group of bits associated and a corresponding set of reliability values Where i is the time index; ii) based on the first transmitted word x i and the set of bits determine (206) the error word e i , iii) based on the set of corresponding reliability values compress (208) the error word e i to generate a first word x i+1 and iv) Transmit (210) the first word x to the second communication device (300) in the first channel (510) i+1 ; wherein, the method (200) further comprises: repeat i) to iv) D times, where i = 0, …, D-1 and D depends on at least one stop condition.

17. A method (400) for a second communication device (300), characterized in that, the method (400) comprises: Receive (402) a first word x from a first communication device (100) in a first channel (510) 0 , wherein the first word x 0 is generated based on an information word m; i) Determine (404) a set of bits i associated with receiving the first word x in the first channel (510) and a corresponding set of reliability values where i is a time index ii) Transmit (406) a second word p to the first communication device (100) in a second channel (520) i , wherein the second word p i indicates the set of bits and the set of corresponding reliability values iii) Receive (408) a first word x from the first communication device (100) in the first channel (510) i+1 ; wherein, the method (400) further comprises: Repeat i) to iii) D times to obtain multiple sets of bits and multiple sets of corresponding reliability values where i = 0, …, D-1 and D depends on at least one stopping condition.

18. A readable storage medium storing a computer program, characterized in that, the computer program has program code for performing the method according to claim 16 or 17 when the computer program is run on a computer.

Images